Short Text Classification Based on Domain Word Ontology
Yu Chuan Wu · 2009
The conventional methods of text classification are not suitable for short text classification because short texts are short and their ability of describing concept is weak.A method using the domain word ontologies for short texts classification was proposed.First,the domain high frequency words were got as the feature words.Then the feature words were extended to concept and sememe by Hownet,which extended the feature from semantic and amended the feature scarcity.Last,the word similarity values were got by calculating the same sememe message included between different concepts.Experimental results prove that the classification efficiency and precision are both improved.